Allogeneic Hematopoietic Stem Cell Transplantation (HCT) Compared to Chemotherapy Only in Acute Myeloid Leukemia (AML) Patients 60 Years and Older: A Center for International Blood and Marrow Transplantation Research (CIBMTR)/ Cancer and Leukemia Group B (CALGB) Study.

Abstract 657

The outcome of patients aged 60 years or older with AML treated with conventional cytotoxic chemotherapy remains extremely poor with few long-term survivors, irrespective of the type of induction or post-remission treatment intensity or duration. In younger patients, allogeneic HCT performed in first complete remission (CR1) appears to offer a potential advantage in terms of progression-free survival and possibly overall survival (OS) in patients with adverse prognostic features, due at least in part to a graft-versus-leukemia (GvL) effect. Preliminary results of reduced-intensity (RIC) allogeneic HCT performed in older AML patients in CR1 suggest that this approach may improve long-term outcome. However, as patients undergoing allogeneic HCT are likely to be highly selected, it remains uncertain if such an approach will truly improve outcome compared to chemotherapy alone. To investigate this possibility, we compared the outcome of 100 AML patients aged 60-70 years who received RIC allogeneic HSCT in CR1 and were reported to the CIBMTR to that of 96 AML patients treated with only standard induction (daunorubicin and cytarabine ± etoposide) and post-remission chemotherapy on CALGB protocols 9720 and 10201 between January 1998 and October 2006. Patients with therapy-related AML or following an antecedent hematologic disorder were included, along with de novo patients. In the chemotherapy-treated group, only patients who remained in CR1 for at least 4 months were included in order to reduce selection bias. Patients in the HCT cohort were younger than those in the chemotherapy-treated cohort (P<.001; median age 63 v 65 years). There was no significant difference in the distribution of sex, therapy-related leukemia, white blood cell count, FAB classification, or the proportion with normal cytogenetics at diagnosis between the two groups. The time from diagnosis to achievement of CR1 was longer for HCT patients compared to chemotherapy-treated patients (P=.007; median 46 v 38 days). HCT donors were HLA-identical siblings (n=48, 48%) and closely matched unrelated donors (n=52, 52%) and all HCT utilized reduced intensity conditioning plus pharmacologic GVHD prophylaxis. The median follow-up of chemotherapy-treated patients was longer than that of HSCT patients (51 v 30 months). Overall, allogeneic HCT tended to be associated with longer leukemia-free survival (LFS) compared to chemotherapy. The 3-year LFS from CR1 for HCT patients was 34% (95% confidence interval [CI], 24%-44%) compared to 17% (95% CI, 10%-25%) for chemotherapy-treated patients (P=.06). This was largely due to a higher relapse rate (RR); 86% of the chemotherapy patients relapsed compared to only 29% who received HCT (P<.001). However, the non-relapse mortality rate was significantly higher among patients receiving HCT compared to chemotherapy-treated patients (P<.001; 39% v 17%). Survival from CR1 did not differ between the groups (P=.47), with 3 year estimates of 35% (95% CI, 25%-46%) for HSCT patients compared to 25% (95% CI, 17%-34%) for chemotherapy patients. Our results indicate that RIC allogeneic HCT in CR1 in older AML patients is associated with a longer LFS, largely due to a reduction in the RR. Strategies aimed at reducing non-relapse mortality associated with allogeneic HCT, including better graft-versus-host disease prophylaxis and treatment, may lead to a significant improvement in OS for older AML patients.